{ "name": "Orchestrator", "nodes": [ { "parameters": { "public": true, "initialMessages": "Give financial metrics for AAPL", "options": {} }, "type": "@n8n/n8n-nodes-langchain.chatTrigger", "typeVersion": 1.3, "position": [ -416, -128 ], "id": "864688c6-1c41-4eab-85ef-8efb51566c9a", "name": "When chat message received" }, { "parameters": { "model": "llama3:latest", "options": {} }, "type": "@n8n/n8n-nodes-langchain.lmChatOllama", "typeVersion": 1, "position": [ -176, 48 ], "id": "29165b1f-3e8d-4c17-acce-32940c94fa71", "name": "Ollama Chat Model" }, { "parameters": { "messages": { "messageValues": [ { "message": "You are a tool dispatcher. Analyze the user's request and return a JSON object with the tool name and parameters. Do not include any explanation or markdown formatting. Return only the JSON object.\n\nAvailable tools and their parameters:\n\n- **patent-monitor**\n - days_back (integer, required): Number of days to look back for patents\n - email (string, optional): Email address for notifications\n\n- **fin-metrics**\n - ticker (string, required): Stock ticker symbol (e.g., \"AAPL\", \"TSLA\", \"MSFT\")\n - narrative (boolean, required): Whether to parse narrative content from SEC filings. Analyze the user's question to determine if they want narrative analysis (mentions of \"insights\", \"analysis\", \"full report\", \"detailed\") or just metrics/numbers.\n\n- **news-sentiment**\n - company (string, required): Company name to search for recent news\n - email (string, optional): Email address to send the sentiment report\n\n- **forecasting**\n - ticker (string, required): Stock ticker symbol to forecast\n - email (string, required): Email address to send forecast results\n - horizon (integer, optional, default: 30): Number of days to forecast\n\nReturn ONLY valid JSON in this exact format with no explanation:\n```json\n{\n \"tool\": \"tool-name\",\n \"params\": {\n \"param1\": \"value1\",\n \"param2\": value2\n }\n}" } ] }, "batching": {} }, "type": "@n8n/n8n-nodes-langchain.chainLlm", "typeVersion": 1.7, "position": [ -80, -176 ], "id": "c1b88e65-54b4-4227-a7ca-bd10a9eeb1a1", "name": "The Brain" }, { "parameters": { "jsCode": "// Get the raw LLM output\nlet raw = $json[\"text\"] || $json[\"data\"] || \"\";\n\n// Clean up formatting issues\nraw = raw\n .replace(/```(json)?/gi, '')\n .replace(/\\\\n/g, '')\n .replace(/\\n/g, '')\n .trim();\n\nlet parsed = {};\ntry {\n parsed = JSON.parse(raw);\n} catch (e) {\n return [{\n json: {\n tool: \"error\",\n params: { message: \"Failed to parse JSON\", raw },\n url: null\n }\n }];\n}\n\n// Define tool configurations with URLs and expected parameter schemas\nconst toolConfigs = {\n \"patent-monitor\": {\n url: \"http://localhost:5678/webhook/patent-monitor\",\n schema: {\n days_back: \"number\", // required\n email: \"string\" // optional\n }\n },\n \"fin-metrics\": {\n url: \"http://localhost:5678/webhook/sec-metrics\",\n schema: {\n ticker: \"string\" ,\n narrative: \"boolean\" // required\n }\n },\n \"news-sentiment\": {\n url: \"http://localhost:5678/webhook/news-sentiment\",\n schema: {\n company: \"string\", // required\n email: \"string\" // optional\n }\n },\n \"forecasting\": {\n url: \"http://localhost:5678/webhook/forecasting\",\n schema: {\n ticker: \"string\", // required\n email: \"string\", // required\n horizon: \"number\"\n }\n }\n};\n\n// Validate and structure parameters\nfunction validateParams(tool, params) {\n const config = toolConfigs[tool];\n if (!config) {\n return { \n valid: false, \n error: `Unknown tool: ${tool}`,\n params: null \n };\n }\n\n const validatedParams = {};\n const schema = config.schema;\n\n // Check each parameter against schema\n for (const [key, value] of Object.entries(params)) {\n if (schema.hasOwnProperty(key)) {\n const expectedType = schema[key];\n const actualType = Array.isArray(value) ? \"array\" : typeof value;\n \n if (actualType === expectedType || (actualType === \"object\" && expectedType === \"object\")) {\n validatedParams[key] = value;\n } else {\n return {\n valid: false,\n error: `Parameter '${key}' should be ${expectedType}, got ${actualType}`,\n params: null\n };\n }\n }\n }\n\n return { valid: true, params: validatedParams };\n}\n\n// Function to build URL with query parameters (manual construction)\nfunction buildUrlWithParams(baseUrl, params) {\n if (!params || Object.keys(params).length === 0) {\n return baseUrl;\n }\n \n // Manually build query string\n const queryParts = [];\n for (const [key, value] of Object.entries(params)) {\n queryParts.push(`${encodeURIComponent(key)}=${encodeURIComponent(value)}`);\n }\n \n const queryString = queryParts.join('&');\n return `${baseUrl}?${queryString}`;\n}\n\n// Validate the parsed tool and params\nconst validation = validateParams(parsed.tool, parsed.params || {});\n\nif (!validation.valid) {\n return [{\n json: {\n tool: \"error\",\n params: { \n message: validation.error,\n original_tool: parsed.tool,\n original_params: parsed.params\n },\n url: null\n }\n }];\n}\n\n// Get base URL and build full URL with query parameters\nconst baseUrl = toolConfigs[parsed.tool].url;\nconst fullUrl = buildUrlWithParams(baseUrl, validation.params);\n\n// Build final payload\nreturn [{\n json: {\n tool: parsed.tool,\n params: validation.params,\n baseUrl: baseUrl,\n url: fullUrl,\n metadata: {\n timestamp: new Date().toISOString(),\n validated: true\n }\n }\n}];" }, "type": "n8n-nodes-base.code", "typeVersion": 2, "position": [ 272, -176 ], "id": "9dc7e552-9bb0-44af-8339-6f719b2b093b", "name": "The Processor" }, { "parameters": { "url": "={{ $json.url }}", "options": {} }, "type": "n8n-nodes-base.httpRequest", "typeVersion": 4.2, "position": [ 480, -176 ], "id": "95408abc-d8fe-46e9-abd0-e3b5fac723f4", "name": "The Router" }, { "id": "b9b4953f-3e1c-5473-b83e-c06bdadb82f4", "name": "Publication ยท Overview and setup", "type": "n8n-nodes-base.stickyNote", "typeVersion": 1, "position": [ -1216, -176 ], "parameters": { "content": "# Orchestrator\n\n## Purpose and structure\nThis folder contains the **Orchestrator** workflow. Its executable nodes use `chainLlm`, `chatTrigger`, `code`, `httpRequest`, `lmChatOllama`. The table and connection map below are derived from the JSON, not an inferred guarantee of behavior.\n\n## Setup\nImport inactive. Configure your own credentials and replace REPLACE_WITH_* values. Review external destinations and Code nodes before running.\n\n## Entry points\nWhen chat message received\n\n## Verification\nSanitized source candidate; not runtime-tested. This overview is generated from node metadata. Read the folder README for the connection map, credential types, and license.\n\n## Attribution\nNik Bear Brown; rights confirmed 2026-10-19\nMIT; preserve source notices.", "width": 700, "height": 800, "color": 1 } } ], "connections": { "Ollama Chat Model": { "ai_languageModel": [ [ { "node": "The Brain", "type": "ai_languageModel", "index": 0 } ] ] }, "When chat message received": { "main": [ [ { "node": "The Brain", "type": "main", "index": 0 } ] ] }, "The Brain": { "main": [ [ { "node": "The Processor", "type": "main", "index": 0 } ] ] }, "The Processor": { "main": [ [ { "node": "The Router", "type": "main", "index": 0 } ] ] } }, "active": false, "settings": { "executionOrder": "v1" } }